Exploring the service map
Each node represents a service, identified by theservice.name resource attribute. Edges (dashed lines) connect services where a client span in one matches a server span in another. Node size reflects relative traffic volume. In Error rate mode, red nodes indicate services with errors in the selected time range. Use the legend in the top-right corner to switch between Latency, Error rate, and Throughput.
The toolbar above the map lets you filter and adjust the view.
Source selector — choose the trace source used to build the map, such as Traces.
Sampling slider — adjust the sampling rate to balance performance and accuracy. Lower rates load faster on high-volume clusters.
Date range picker — set the time window for the trace data used to build the map.
Use the +/- buttons in the bottom-left corner of the map or scroll to zoom in and out.